Kinematic and Dynamic Analyses of the Orthoglide 5-axis

bistredingdongMechanics

Oct 31, 2013 (3 years and 7 months ago)

114 views






Abstract

This paper deals with the kinematic and dynamic
analyses of the Orthoglide 5
-
axis, a five
-
degree
-
of
-
freedom
manipulator. It is derived from two manipulators: i)

the
Orthoglide

3
-
axis; a three dof translational manipulator and
ii) the Agile eye;
a parallel spherical wrist. First, the
kinematic and dynamic models of the Orthoglide 5
-
axis are
developed. The geometric and inertial parameters of the
manipulator are determined by means of a CAD software.
Then, the required motors performances are evalu
ated for
some test trajectories. Finally, the motors are selected in the
catalogue from the previous results.

I.

I
NTRODUCTION


ARALLEL kinematics machines become more and more
popular in industrial applications [
1
,

2
].This growing
attention is inspired by their essential advantages over serial
manipulators that have already reached the dynamic
performance limits. In contrast, parallel manipulators are
claimed to offer better ac
curacy, lower mass/inertia
properties, and higher structural stiffness (i.e. stiffness
-
to
-
mass ratio)

[
3
]. These features are induced by their specific
kinematic structure, which resists to the error accumulation in

kinematic chains and allows convenient actuators location
close the manipulator base. Besides, the links work in parallel
against the external force/torque, eliminating the cantilever
-
type loading and increasing the manipulator stiffness [
4
].
The latter makes them attractive for innovative machine
-
tool
architectures [
5
,
6
], but practical utilization for the potential
benefits requires devel
opment of efficient kinematic and
dynamic analyses, which satisfy the computational speed
and accuracy requirements of relevant design procedures.

T
his paper focuses on the

kinematic and dynamic
analyses of the Orthoglide 5
-
axis, a spatial parallel
-
kinema
tics
machine (PKM) developed for high speed operations [
7
]. To
evaluate the forces and torques that have to be exerted by
the actuators of the Orthoglide 5
-
axis, its kinematic and
dynamic analyses are of prim
ary importance.

II.

O
RTHOGLIDE
5
-
AXIS

The Orthoglide 5
-
axis, illustrated in
Fig.
1
, is derived from a
3
-
dof translating manipulator, the Orthoglide 3
-
axis and a 2
-
dof spherical wrist [
7
].




Fig.
1
. The Orthoglide 5
-
axis

The Orthoglide 3
-
axis is a Delta
-
type PKM [
8
] dedicated to
3
-
axis rapid machining applications developed at the
Research Institute in Communica
tions and Cybernetics of
Nantes (IRCCyN) [
9
]. This mechanism is composed of three
identical legs. Each leg is made up of a prismatic joint, a
revolute joint, a parallelogram joint and another revolute
joint. The fir
st joint, i.e. the prismatic joint of each leg, is
actuated while the end
-
effector is attached to the other end
of each leg. Hence, the Orthoglide 3
-
axis is a PKM with
movable foot points and constant chain lengths.


Fig.
2
. The Ort
hoglide 3
-
axis

The Orthoglide 3
-
axis gathers the advantages of both
serial and parallel kinematic architectures such as regular
workspace, homogeneous performances, good dynamic
performances and stiffness. The interesting features of the
Orthoglide 3
-
axis
are large regular dextrous workspace,
uniform kinetostatic performances, good compactness [
10
]
and high stiffness [
11
].

Kinematic and Dynamic Analyses of the Orthoglide 5
-
axis

R. Ur
-
Rehman, S. Caro, D. Chablat, P. Wenger

Institut de Recherche en Communication et Cybernétique de Nantes,

UMR CNRS 6597, 1 rue de la Noë, 44321 Nantes, France

Raza.Ur
-
Rehman@irccyn.ec
-
nantes.f
r

P




The two
-
dof spherical wrist that is implemented
in the
Orthoglide 5
-
axis is derived from the Agile Eye, a three
-
dof
spherical wrist developed by Gosselin and Hamel

[
12
].
Nevertheless, the two
-
dof spherical wrist was designed in
order to have high stiffness, [
13
]. A CAD model of the wrist
is shown in
Fig.
3
. It consists of a closed kinematic chain
composed of five components: the proximal 1, the proximal 2,
the distal, the terminal
and the base. These five links are
connected by means of revolute joints, the two revolute
joints connected to the base being actuated. Let us notice
that the revolute joints axes intersect.


Fig.
3
. Spherical wrist of the Orthoglid
e 5
-
axis

III.

T
RAJECTORY
P
LANNING

In order to analyze the kinematic and dynamic performance

of the Orthoglide 5
-
axis, two test trajectories are proposed.
The inverse kinematic and dynamic problems of the
manipulator are solved for both trajectories.

The Orthogl
ide 5
-
axis is designed for a 500x500x500

mm
3
cubic workspace. Let us notice that the cubic workspace
center, i.e., point

I

, and the origin
O

of the reference frame,
and the intersection of the prismatic joints axes, do not
coincide as explained in[
9
] and shown in
Fig.

4
. The
vector
, expressed by
dr

=

[
dx dy dz
]
T
, is also called
position vector of the cubic workspace center. Here, the
geometric cent
re of the path of the test trajectories is
supposed to be point
I
. Accordingly, the test trajectories are
defined as follows:



Traj.

I
: semi
-
circular trajectory in a plane perpendicular to
XY
-
plane defined by radius
R

, trajectory angle ψ with
Y
-
axis, traj
ectory plane orientation angle φ (angle between
the trajectory plane and X
-
axis) and vector
v

orientation
angle δ
(
Fig.

4
). Position vector
p

and wrist orientation
vector
v

are given by:



w
here δ varies from π/6 to 5π/6 and ψ varies from 0 to π..



Traj.

II
: circular trajectory in horizontal or
XY
-
plane
defined by radius
R
, vector
v,

constant orientation angle
γ with
Z
-
axis and the angle
ψ (
Fig.
5
). Position vector
p

and wrist orientation vector
v

are given by:



where ψ varies from 0 to 2π


Fig.

4
. Orientation of vector
v

(Traj
I
)


Fig.
5
. Orientation of vector
v

(Traj
II
)

Let us notice that several test trajectories can be derived
from
Traj.

I

and Traj.

II
, by changing their parameters.




IV.

K
INEMATIC
A
NALYSIS

A.

Orthoglide 3 axis

The geometric parameters of the Orthoglide 3
-
axi
s are
defined as a function of the size of a prescribed cubic
Cartesian workspace, that is free of singularities and internal
collision

The kinematic architecture of the Orthoglide
-
3axis is
shown in
Fig.
2

where
A
1
B
1
,
A
2
B
2

and
A
3
B
3

represent the
prismatic joints and
P

is the end
-
effector. Due to its Delta
-
linear architecture, the Orthoglide
-
3axis is a translating
parallel manipulator with 3
-
DOF.

A simplified model of the Orthoglide 3
-
axis is illustrated in
Fig.
6

[
14
] in which three links of length
L

are connected by
means of a spherical joint to end
-
effector “
P
” at one end and
to the corresponding prismatic joints “
A
i
” at the othe
r end.
θ
x
, θ
y

and θ
z

are the angles between the links and the
corresponding prismatic joints axes. The reference frame is
coincident with the prismatic joint axes; it is the origin being
the intersection point of those axes. The input position
vector of th
e prismatic joint variables is represented by

and the output position vector of the end
-
effector by
.


Fig.
6
. Simplified model of the Orthoglide 3
-
axis

Using these notations,
the inverse kinematic relations for a

spherical singularity free workspace can be written as [
10
]


Due to the Orthoglide geometry and manufacturing
technology, the displacement of its p
rismatic joints is
bounded

[
10
], namely,


The kinematic performance of the Orthoglide 3
-
axis is
analyzed by means of the foregoing test trajectories. The

velocity of end
-
effector
P

thro
ughout the trajectory, is
supposed to be constant i.e. V
p
=

1 m/s. Accordingly the
actuated prismatic joints position, rates and acceleration are
plotted in Figs. 7 to 9 for both test trajectories, the radius of
their path being equal to 0.2

m.


Fig.
7
. Actuated prismatic joints position, rates and acceleration of the
Orthoglide 3
-
axis for Traj

I

with φ=90˚


Fig.
8
. Actuated prismatic joints position, rates and acceleration of the
Orthoglide 3
-
axis for Traj

I

w
ith φ=45˚





Fig.
9
. Actuated prismatic joints position, rates and acceleration of the
Orthoglide 3
-
axis for Traj

II

with

γ=45˚

Fig.
7

shows the kinematic performance required by the
prismati
c actuators when end
-
effector
P

moves in
YZ
-
plane.
Even if
P

does not move along
X
-
axis, the displacement of
the prismatic actuator mounted along
X
-
axis is not null.
Fig.
8

and
Fig.

9

display the required kinematic performance of the
motors when
P

follows Traj

I

(
φ=45˚) and Traj

II (
γ=45˚
)
,
respectively for
R
=0.2m. The maximum velocities and
accelerations of the prismatic actuators required for the three
test trajectories are shown in
Table
I
. We can notice that the
maximum

prismatic joint velocity is equal to 1 m/s whereas its
maximum acceleration is equal to 6.31 m/s
2
.

T
ABLE
I
.

M
AXIMUM PRISMATIC JOI
NTS RATES AND ACCELE
R
A
TIONS

Test
Trajectory

Max Absolute
Velocity [m/s]

Max Absolute
Acceleration [m/
s
2
]


V
x

V
y

V
z

A
x

A
y

A
z

Traj
I
, γ
=90˚

0.12

1.01

1.06

0.62

6.32

6.32

Traj
I
, γ
=45˚

0.77

0.77

1.07

4.51

4.51

6.33

Traj
II
, γ
=45˚

1.06

1.06

0.12

6.31

6.31

0.66

B.

Spherical Wrist

The spherical wrist mechanism of the Orthoglide 5
-
axis
consists of a closed kinematic chain composed of five
components: proximal
-
1, proximal
-
2, distal, terminal and the
base. These five links are connected by means of revolute
joints, of which axes intersect. Besides, only the two revolute
joints connected to the base of the wrist are actuated. The
distal has an

imaginary axis of rotation passing through the
intersection point of other joint axis and perpendicular to the
plane of proximal
-
2.

The kinematic equations of the wrist are written by means
of six reference frames attached to the six rigid bodies and
the
corresponding Denavit
-
Hartenberg (DH)
-
parameters.
Fig.
10

shows the orientation of these reference frames while
vector
v

represents the orientation of the terminal of the
wrist, i.e. the cutting tool. Moreover, α
0

denotes the angle
between
e
1

and
e
2

while α
i

denotes the angle between
e
i

and
e
i+2

(
i

=

1…4).


Fig.
10
. Orientations of reference frames for the Orthoglide Wrist

The reference frame
R
1

is defined in such a way that the
Z
1
-
axis coin
cides with
e
1

and
e
2

lies in the X
1
Z
1
-
plane.
Similarly
R
2

has its Z
2
-
axis in the direction of
e
2

and
e
1

lies in
the X
2
Z
2
-
plane. Reference frame
R
i

(
i

=

3,

4,

5,

6) with Z
i

=

e
i

are defined by the rotation of frame R
i
-
2

and following the DH
conventions.

Fin
ally, the inverse kinematic problem of the wrist can be
derived from the definitions of the reference frames and unit
vectors to develop the relations between the joints angles
(
θ
1
,

θ
2
,

θ
3
,

θ
4
) [
15
].
Fig.
11

to
Fig.

13

display the revolute
joints angles, rates and accelerations for the two trajectories
introdu
ced in Section

III, the radius of their path being equal
to 0.2 m.



Fig.
11
. Revolute joints angles, rates and accelerations of the wrist
(Traj

I
,

φ=90˚)


Fig.
12
. Revolute joints angles, rates and accelera
tions of the wrist
(Traj

I
,

φ=45˚)

Fig.
11

is the case where wrist end
-
effector moves in the
YZ
-
plane (φ=90˚) so only one of the wrist actuator (θ
1
) works,
while
Fig.
12

a
nd
Fig.

13

represent the cases where both of
the actuators work. Compared to Traj
I
, both wrist actuators
experience greater velocities and accelerations for Traj
II

(max

velocity=5 m/s and max acceleration=22 m/s
2
). This can be
explained by the higher order variations of rotation angles in
Traj
II

to that of linear variation of rotation angles in Traj
I





Fig.
13
. Revolute joints angles, rates and accelerations of the wrist
(Traj

II
,

γ=45˚)

V.

D
YNAMIC
A
NALYSIS

A.

Orthoglide 3
-
axis

The dynamic analysis of the Orthoglide 3
-
axis is performed

in order to evaluate the torques required by the three
actuated prismatic joints. Here, we take advantage of the
dynamic model developed in

[
16
]. The geometric and
dynamic parameters used in the analysis are obtained from
SYMORO+ (SYmbolic MOdeling of Robots), a software for
the automatic generation of symbolic model of robots [
17
]
and d
efined in

[
16
].
Fig.
14

illustrates a leg of the Orthoglide
with the definition of the parameters and the frames attached
to all the bodies

[
16
].


Fig.
14
. Orthoglide leg parameterization for the dynamic analysis

The geometric, mass and inertial parameters used in the
dynamic model were determined for the Orthoglide 5
-
axis by
means of SolidWorks CAD software, and
from the geometry
of the mechanism. The dynamic performance of the
Orthoglide 3
-
axis is then evaluated for different test
trajectories. The actuators forces required to follow those
trajectories are shown in
Fig.
15
.

T
ABLE
II
.

PARAMETERS OF
O
RTHOGLIDE
3
-
AXIS REQUIRED FOR DY
NAMIC
ANALYSIS

Parameters for leg

i

Symbol

Mass of the platform (wrist)

m
p

Mass of body 1

m
1
i

Mass of bodies 2 and 4

m
2
i
,
m
4
i

Mass of bodies 3 and 7

m
3
i
,
m
7
i

Length of
parallelograms

d
4
i

Width of parallelograms

2

r
2
i

Actuators moment of inertia

I
mi


Fig.
15
. Orthoglide 3
-
axis actuators forces (Traj
-
I

and
II
)

B.

Spherical Wrist

A Newton approach is used to come up with the dynamic
modeling of the
Orthoglide wrist. A similar methodology was
also used in [
15
]. Let us assume that:



friction forces are neglected;



there is a spherical joint is between the distal and the
terminal link in order to get an isostatic mec
hanism;



there is a planar joint between the distal and the
proximal
-
2.

Thus, the free body diagrams of terminal, distal, proximal
-
1
and proximal
-
2 can be drawn. The equilibrium equations are
written for each free body diagram and then, the equations
used t
o evaluate the actuators torques are obtained

[
15
]. The
latter are shown in
Fig.
16
. It can be seen that for
YZ
-
plane
trajectory (Traj
I
, φ

=

90˚), only the first actuator, al
igned with
the
Y
-
axis (
Fig.

4
) experiences the torque, while for other two
test trajectories both actuators work and experience torques.
Second actuator, aligned with
X
-
axis, experiences greater
torque compared to
the first actuator for Traj
I
, φ

=

45˚and for
Traj
II





In order to verify the results obtained with the Newton
approach, the principle of virtual work is used.

As a matter of fact, variations in kinetic and potential
energies i.e. Δ
KE

and Δ
PE

are evaluated during a time
interval
dt
.
The total energy variation
ΔE

over
dt

is defined
as ΔE

=

ΔKE

+

ΔPE. Therefore, the total virtual work
W

is
calculated by the product of the mean torques and the
corresponding angular displacement during each time
interval
dt
. i.e.,


and
being the mean torques of actuators 1 and 2,
respectively during
dt, respectively
.


Fig.
16
. Spherical wrist actuators torques (Traj
-
I

and
II
)

It is noteworthy that the
difference between the global
virtual work
W

and
ΔE

should be null due to energy
conservation. Accordingly, we compute the difference
between
W

and
ΔE

and check the dynamic model of the wrist,

i.e., ΔW= ΔE
-
W. This difference is highlighted in
Fig.
17

for
the thre
e test trajectories that are considered in the scope of
the study. It turns out ΔW is null in all cases. Consequently,
the dynamic model of the wrist makes sense.


Fig.
17
. Energy balance for wrist dynamics (Traj
-
I and II)

VI.

C
ONCLUSIO
N

This paper dealt with the kinematic and dynamic analyses
of the Orthoglide 5
-
axis, a five
-
degree
-
of
-
freedom
manipulator.

First, it turned out that kinematic and dynamic
analyses of the translating part and the spherical wrist of the
manipulators can be
decoupled. The geometric and inertial
parameters of the manipulator were determined by means of a
CAD software. We came up with the dynamic model of the
spherical wrist by means of a Newton approach. Besides, this

model has been checked with the principle
of virtual work.
Then, the required motors performances were evaluated for
some test trajectories.
Various simulations results showed
that the FFA 20
-
80 harmonic drive motors of 0.8 kW and the
NX430 EAF motors of 1.8 kW, primarily selected for the wrist
an
d Orthoglide 3
-
axis respectively, are suitable for the
prototype of the Orthoglide 5
-
axis.
In future works, friction
forces as well as payload will be considered in the dynamic
analysis and further test trajectories will be performed.

R
EFERENCES

[
1
]

T.Brogardh, “Present and future robot control development
-

An
industrial perspective,”
Annual Reviews in Control
, 31(1), pp. 69
-
79, 2007

[
2
]

H. Chanal, E. Duc, and P. Ray, “A study of the impact of machine
tool structure on machini
ng processes”.
International Journal of
Machine Tools and Manufacture
, 46(2), pp. 98
-
106, 2006.

[
3
]

J.
-
P. Merlet, Parallel Robots, 2nd Edition.
Springer, Heidelberg,
2005.

[
4
]

L. W. Tsai, Robot analysis: The Mechanics of Seria
l and Parallel
Manipulators, John Wiley & Sons, New York, 1999.

[
5
]

J. Tlusty, J. C. Ziegert, and S. Ridgeway, “Fundamental
comparison of the use of serial and parallel kinematics for
machine tools,”
CIRP Annals,

48(1), pp. 351

356, 1999

[
6
]

O. Company, and F. Pierrot, “Modelling and Design Issues of a 3
-
axis Parallel Machine
-
Tool,”
Mechanism and Machine Theory
,
37(11), pp 1325
-
1345, 2002

[
7
]

D. Chablat et P. Wenger, “Device for the movement and
orientation of an

object in space and use thereof in rapid
machining”, Centre National de la Recherche Scientifique /Ecole
Centrale de Nantes, European patent: EP1597017, US patent:
20070062321, 2005.

[
8
]

R. Clavel, "Delta, a fast robot with parallel geometr
y",

Proc. 18
th

International Symposium on Industrial Robots,

1988, pp.91
-
100.

[
9
]

D. Chablat, and P. Wenger, “Architecture Optimization, of a 3
-
DOF Parallel Mechanism for Machining Applications, The
Orthoglide,” IEEE Trans. on Robotics and A
utomation 19, 2003,
pp.403

410.

[
10
]

A. Pashkevich, P. Wenger and D.Chablat, “Design St rat egies for
t he Geomet ric Synt hesis of Ort hoglide
-
t ype Mechanisms”, Journal
of Mechanism and Machine Theory, Vol. 40(8), 2005, pp. 907
-
930.

[
11
]

A. Pashkevich, D. Chablat and P. Wenger, “St iffness Analysis of
3
-
d.o.f. Overconst rained Translat ional Parallel Manipulat ors”,
Proc. IEEE Int. Conf. Rob. and Aut omat ion, May 2008.

[
12
]

C. Gosselin, and J.F. Hamel, “The agile eye: a high
performance
t hree
-
degree
-
of
-
freedom camera
-
orienting device,”
IEEE Int.
conference on Robotics and Automation
, San Diego, 1994,
pp.

781
-
787.

[
13
]

D. Chablat, and P. Wenger, “A six degree of freedom Hapt ic
device based on t he Ort hoglide and a
hybrid Agile Eye,” 30t h



Mechanisms & Robotics Conference (MR), Philadelphia, USA,
2006.

[
14
]

A. Pashkevich, D. Chablat and P. Wenger, “Kinemat ics and
Workspace Analysis of a Three
-
Axis Parallel Manipulat or: t he
Ort hoglide,” Robot ica, Vol 24
,


Issue 1, 2006, Pages: 39
-

49


[
15
]

F. Caron, “Analyse et concept ion d'un manipulat eur parallèle
sphérique à deux degrés de libert é pour l'orient at ion d'une caméra,”
M.S. t hesis, Laval Universit y, Quebec, QC, Canada, 1997.

[[
16
]

S. Guégan, "
Contribution à la modélisation et l'identification
dynamique des robots parallèles
."
Ph.D. Thesis, Ecole Centrale
de Nantes, Nantes, France, 2003

[[
17
]

W. Khalil and D. Creusot , "SYMORO+: a system for the
symbolic modelling
of robots", Robotica, Vol. 15, pp. 153
-
161,
1997